DESCRIPTION: Mortality rises with age, reflecting the progression of physiological deterioration. The pattern of this age-related mortality increase may vary among causes of death, among species, between sexes, and among groups with different characteristics. These variations are important clues on the underlying mechanisms of senescence, evolutionary backgrounds of the mechanisms, relationships between senescence and diseases, and environmental and genetic effects on senescence. A statistical tool for examining the mortality trajectories is the LAR, which is the age-specific rate of relative mortality increase with age and is useful in detecting mortality accelerations and decelerations. Previous studies, however, did not analyze the differential age trajectories of mortality in depth. This is due to the prevalent view that the age patterns of mortality over a wide range of adult ages (except for very old ages) are similar for major degenerative causes of death and in many human and nonhuman populations, since they fit the exponential function fairly well. Thus those studies usually examined the average LAR (i.e., the Gompertzian aging rate) over age, but did not focus on age variations in the LAR. Serious limitations of this conventional approach have been revealed in some recent studies, which indicate significant age variations of the LAR not only at oldest ages but also at middle ages and younger old ages, and substantial differences in the LAR pattern between sexes, among diseases, and over time. In this project, the LAR analysis will be used for investigating (1) inter-species differentials in the age pattern of mortality and (2) environmental and genetic effects on the age pattern of mortality in humans and laboratory animals. In particular, the following assumptions and hypotheses will be tested: (a) the age-related mortality increase tends to slow down in earlier life stages in r-strategy (high fertility) species than in K-strategy (low mortality) species; (b) the mortality deceleration is more likely to be preceded by a phase of mortality acceleration in vertebrates than in invert-ebrates; (c) effects of factors that extend the lives of laboratory animals differ with respect to four major dimensions (initial mortality, oldest-old mortality, pace of senescence, and timing of senescence); (d) risk factors involved in the accumulation of free radical damages in humans (e.g., smoking and diet) affect the rate of age-associated mortality increase; and (e) the diffusion of smoking has altered the age pattern of male mortality consid-erably. A new multivariate method, called the power-function hazard model analysis, will be developed for analyzing effects of risk factors on aging rates.